The phone rings during the rush
Between orders, the pass and guests, nobody can answer every call. Without a structured callback, names, party sizes, dates and special requests are easily incomplete.
Unanswered enquiry
More guests, fewer interruptions
AlpenAgent connects phone, website and existing workflows into one clear guest service. Reservation requests are captured completely, menu questions are answered from approved information and sensitive cases are handed to people.
Where revenue disappears during service
Not every lost reservation looks like an error. The phone rings at the wrong moment, a question remains unanswered or a guest never receives a clear confirmation. We make these transitions visible and manageable.
No-shows without clear confirmation
A reservation becomes reliable only when details, capacity and confirmation agree. Reminders can help, but they do not replace clear rules for changes and cancellations.
Empty table at peak time
Feedback remains unanswered
Reviews and complaints often arrive after closing. A prepared draft saves time while publication, compensation and critical replies remain with the team.
Reputation needs ownership
Three working areas, one clear guest contact
The solution does not sit beside your operation. It follows existing routes and may only take actions that have been agreed in advance.
Voice booking
The phone assistant captures recurring reservation requests, asks for date, time, party size, name and relevant notes, then sends the request to an approved reservation system or the team. If availability is unclear, it does not promise a table.
Menu and service knowledge
On the website or phone, the assistant answers from approved information: current menu, allergens, opening hours, directions, group policies or takeaway. Contradictory or individual dietary questions are handed over.
Review and callback process
After a visit, feedback can be structured, draft replies prepared and callbacks triggered. Critical reviews, compensation and public responses remain subject to human approval.
From the first ring to a clean handover
A useful workflow does not begin with a bot. It begins with the rules of your restaurant: which details are mandatory, when a table is truly available and who handles exceptions.
Understand the request
The guest names a date, party size and preferred time. The assistant recognises the purpose and asks only for missing information.
Check capacity and rules
Approved opening hours, blocked periods, group sizes and the connected reservation logic determine which next step is allowed.
Confirm or hand over
A booking is confirmed only when the integration permits it. Otherwise the team receives a complete request instead of an ambiguous note.
Keep the context
Changes, cancellations, callbacks and reviews remain traceable. The team can see what was asked, promised and left open.
The same restaurant with a calmer information flow
during lunch and evening service, calls, paper notes and follow-up questions move between several people. Guests repeat details, staff search for information and nobody is certain whether a request has already been confirmed.
recurring questions are answered from approved knowledge. Reservation requests contain the required details and arrive at the agreed destination. Exceptions and sensitive decisions reach a person together with the conversation context.
What missed reservation enquiries may cost
This calculator is a transparent illustration, not a revenue guarantee. It multiplies missed enquiries, average guest value and an assumed conversion rate. Adjust every value to your own operation.
Potential revenue missed per monthCHF 3'000
Illustrative recoverable revenueCHF 2'700
Projected over one year
What to clarify before a restaurant pilot
Does the assistant book tables autonomously?
Only within agreed rules and only when the connected reservation logic permits confirmation. Otherwise it collects the request and hands it to the team.
Will it work with our existing table plan?
That depends on interfaces and daily routines. We review the reservation system, capacity rules, blocked periods and required actions before a pilot. Replacing the system is not automatically necessary.
How does it answer allergen questions?
It uses only approved, maintained information. Individual or medically sensitive questions are never guessed and are handed to a responsible person.
Can it work in German, French, Italian and English?
Yes. Dialogues can be designed in several languages. Before launch we define and test language, tone, follow-up questions and handovers for the languages actually required.
How quickly can a pilot start?
A limited use case can often be prepared within a few weeks. Timing depends on data quality, integrations, approvals and the required test scope.
How is success measured?
Useful measures include complete enquiries, avoided abandonments, human handovers, confirmed reservations and time saved on recurring guest questions.
Relevant building blocks for your operation
This industry page explains the workflow. These service pages describe the technical implementation in detail.
Start with one real service moment
Bring ten typical guest questions, your reservation process and the most important exceptions. In the Business Check we identify a limited use case that can create measurable relief and what needs to be organised first.
Start the Business Check